A systems approach to inflammation identifies therapeutic targets in SARSCoV-2 infection
clinical immunology/immunity inflammation proteomics
Authors: Frank L. van de Veerdonk et al.,
Journal/ Pre-Print: MedRxiv
Tags: Immunology/Immunity; Inflammation; Proteomics
1. Proteomics analysis of plasma from 119 patients with COVID-19 comparing patients admitted to the intensive care unit (ICU) versus clinical ward. Results indicated upregulation of IL-1/IL-6 pathway in severely ill ICU patients.
2. Patients with severe COVID-19 showed increased expression of inhibitory enzymes for Kinin-Kallikrein system (Serpina5, Serpina12); activated kinin-kallikrein system may offer a mechanism for pulmonary angioedema seen in COVID-19.
3. Patients with severe COVID-19 showed decrease in stem cell factor (SCF) suggesting a role in disruption of haematopoiesis and lymphopenia seen in COVID-19
The authors performed a prospective observational study of all consenting patients with COVID-19 at a tertiary care centre in the Netherlands. They compared the plasma proteome profile of patients with or without admission to intensive care unit (ICU). They also used two cohorts of patients as controls: healthy controls and patients with sepsis. The proteomics analysis included commercial ELISA assays for four proteins, IL6, TNF-alpha, C3b and terminal complement complex, and 3 Olink® platforms (Inflammation, Cardiometabolic and Cardiovascular II) testing a total of 267 proteins. Statistical analysis was performed using the Man-Whitney test or the Kruskal-Wallis test with Dunn’s multiple comparison test. Results demonstrated significant differences between the ICU vs non-ICU groups, with increase in IL-1/IL-6 pathway, complement system and kinin-kallikrein pathways in ICU patients. Unsupervised clustering suggested possible homogeneous pathogenetic mechanism in most patients, with patients clustering based largely on disease severity, and not based on any obvious inflammatory endotypes.
Impact for SARS-CoV2/COVID19 research efforts
Understand the immune response to SARS-CoV2/COVID19
· Clinical study
Strengths and limitations of the paper
Novelty: Use of two orthologous techniques (ELISA and the Olink) to characterise large profile of plasma proteins associated with severity in COVID-19.
Standing in the field: The findings are consistent with previous work, and strengthen out understand of the host response in COVID-19. The study confirms the strong inflammatory role of IL-6 in COVID-19 disease severity but expands this by further cytokines and novel pathways (kinin-kallikrein).
Appropriate statistics: Mostly. There was no reported sample size calculation. The authors report the use of the Mann-Whitney for comparing two groups, and the Kruskal-Wallis + Dunn’s test for multiple groups. However, it is not clear why a uniform approach was not taken. Additionally, a correction for multiple testing (a good idea) is reported in one of the figures, but not in the method, so unclear when it was applied. Lastly, as some of the results were very positive, it would have been better to perform a roc analysis for the proteins identified as potential biomarkers, such that a cut-off could be made, and this would have contributed to the translatability (e.g. use as biomarkers) of the results.
Viral model used: Hospitalized patients with confirmed or assumed COVID-19 infection (ICU- & non-ICU patients)
Translatability: Yes, although there is limited discussion of the translatability in the discussion, it is expected that as some of the proteins identified as biomarkers of severe COVID-19 are detectable by commercial ELISAs, this could relatively easily be translated to clinical testing.
Supports trials of anti-IL-6 receptor antibodies and IL-1 receptor antagonist in treating COVID-19 and inhibiting the kinin-kallikrein system might serve as another potential treatment for severe-stage COVID-19 patients.
1. The study is reported a multicentre study, but recruitment appears to have been performed in a single centre.
2. COVID-19 confirmation was based on 1) detection of SARS-CoV-2 or 2) presumed infection due to symptoms, signs and tomography scans. The latter is reasonable, but it would have been useful to report the definition for tomography confirmation.
3. The proteomics analysis is targeted, rather than unbiased/ discovery, limiting the number of potential biomarkers to the ELISA/ Olink panels available.
4. The controls are reasonable, although the demographics are not reported to assess the comparability. It would have been better to have included controls sampled from the same hospital in the same way, at the same time, ideally to have included patients with suspected COVID-19 that were not COVID-19 positive.
5. It is reported that blood samples were taken frequently during the admission, but unclear which samples were used for the proteomic analysis. This is very important for the translatability of the results.
6. Statistical analysis comments as above.
7. Study searches for immune endotypes but biomarkers with detection in <80% of samples were excluded for proteomic analysis possible omitting smaller groups of immune endotypes.